Project Summary Designing enzymes that are as efficient as natural enzymes is very difficult, showing that we understand too little about the biophysical constraints that limit evolutionary routes to new functions. In-depth studies about the biophysical constraints on protein evolution have been limited to a few model systems. These studies indicate that epistasis, which occurs when mutations have different effects in different sequence contexts, is common, but its biophysical basis and pervasiveness are not well studied. Promiscuity, which is the coincidental ability to carry out a reaction that is not a biological function, also plays a role by providing the raw material for natural selection to evolve new biological functions. Underlying promiscuity and epistasis is protein biophysics: structure, stability, dynamics, and enzymatic mechanism. The goal of this proposal is to illuminate the roles of promiscuity and epistasis in protein evolution by comparing biophysical constraints on promiscuous enzymes to those of highly specific enzymes. We established the N-succinylamino acid racemase (NSAR)/o-succinylbenzoate synthase (OSBS) subfamily as one of the best models for determining the role of promiscuity in enzyme evolution. NSAR activity evolved from an ancestral OSBS, and many enzymes are catalytically promiscuous for both activities. Those that have NSAR activity also exhibit substrate promiscuity, preferring hydrophobic N-succinylamino acids but having weak activity with other side chains. This proposal hypothesizes that promiscuity is correlated with the nature and extent of biophysical constraints on mutations that are required to evolve new activities. Our aims are to 1) Define the biophysical constraints on evolution of NSAR activity in a highly specific OSBS enzyme. We will test the hypothesis that several highly specific OSBSs will evolve NSAR activity by different routes due to epistasis and other biophysical constraints; 2) Compare biophysical constraints on promiscuous NSAR/OSBS enzymes and highly specific OSBS enzymes by testing the hypothesis that changing the N-succinylamino acid preference of promiscuous NSAR/OSBS enzymes will be more feasible than changing substrate preference of highly specific OSBSs; and 3) Develop a general approach to identify epistatic interactions. These aims will be a significant step toward determining how structure, stability, dynamics and catalytic mechanism affect the evolution of new enzyme activities. Comparing promiscuous and highly specific enzymes will clarify the role of promiscuity. Biophysical analysis will reveal epistatic mechanisms, especially effects of mutations distant from the active site. We will leverage this data with a massively parallel analysis of mutation phenotypes to develop a general approach to identify epistatic interactions. We will use this approach to refine protein engineering strategies, steering mutagenesis toward epistatic sites that need to be simultaneously optimized.